Spatio-temporal diffusion of dynamic PET images

IE-CycleGAN: improved cycle consistent adversarial network for unpaired PET image enhancement

PURPOSE

Technological advances in instruments have greatly promoted the development of positron emission tomography (PET) scanners. State-of-the-art PET scanners such as uEXPLORER can collect PET images of significantly higher quality. However, these scanners are not currently available in most local hospitals due to the high cost of manufacturing and maintenance. Our study aims to convert low-quality PET images acquired by common PET scanners into images of comparable quality to those obtained by state-of-the-art scanners without the need for paired low- and high-quality PET images.

METHODS

In this paper, we proposed an improved CycleGAN (IE-CycleGAN) model for unpaired PET image enhancement. The proposed method is based on CycleGAN, and the correlation coefficient loss and patient-specific prior loss were added to constrain the structure of the generated images. Furthermore, we defined a normalX-to-advanced training strategy to enhance the generalization ability of the network. The proposed method was validated on unpaired uEXPLORER datasets and Biograph Vision local hospital datasets.

RESULTS

For the uEXPLORER dataset, the proposed method achieved better results than non-local mean filtering (NLM), block-matching and 3D filtering (BM3D), and deep image prior (DIP), which are comparable to Unet (supervised) and CycleGAN (supervised). For the Biograph Vision local hospital datasets, the proposed method achieved higher contrast-to-noise ratios (CNR) and tumor-to-background SUVmax ratios (TBR) than NLM, BM3D, and DIP. In addition, the proposed method showed higher contrast, SUVmax, and TBR than Unet (supervised) and CycleGAN (supervised) when applied to images from different scanners.

CONCLUSION

The proposed unpaired PET image enhancement method outperforms NLM, BM3D, and DIP. Moreover, it performs better than the Unet (supervised) and CycleGAN (supervised) when implemented on local hospital datasets, which demonstrates its excellent generalization ability.

Dose reduction in dynamic synaptic vesicle glycoprotein 2A PET imaging using artificial neural networks

Objective. Reducing dose in positron emission tomography (PET) imaging increases noise in reconstructed dynamic frames, which inevitably results in higher noise and possible bias in subsequently estimated images of kinetic parameters than those estimated in the standard dose case. We report the development of a spatiotemporal denoising technique for reduced-count dynamic frames through integrating a cascade artificial neural network (ANN) with the highly constrained back-projection (HYPR) scheme to improve low-dose parametric imaging.Approach. We implemented and assessed the proposed method using imaging data acquired with11C-UCB-J, a PET radioligand bound to synaptic vesicle glycoprotein 2A (SV2A) in the human brain. The patch-based ANN was trained with a reduced-count frame and its full-count correspondence of a subject and was used in cascade to process dynamic frames of other subjects to further take advantage of its denoising capability. The HYPR strategy was then applied to the spatial ANN processed image frames to make use of the temporal information from the entire dynamic scan.Main results. In all the testing subjects including healthy volunteers and Parkinson's disease patients, the proposed method reduced more noise while introducing minimal bias in dynamic frames and the resulting parametric images, as compared with conventional denoising methods.Significance. Achieving 80% noise reduction with a bias of -2% in dynamic frames, which translates into 75% and 70% of noise reduction in the tracer uptake (bias, -2%) and distribution volume (bias, -5%) images, the proposed ANN+HYPR technique demonstrates the denoising capability equivalent to a 11-fold dose increase for dynamic SV2A PET imaging with11C-UCB-J.

Guided block matching and 4-D transform domain filter projection denoising method for dynamic PET image reconstruction

PURPOSE

Dynamic PET is an essential tool in oncology due to its ability to visualize and quantify radiotracer uptake, which has the potential to improve imaging quality. However, image noise caused by a low photon count in dynamic PET is more significant than in static PET. This study aims to develop a novel denoising method, namely the Guided Block Matching and 4-D Transform Domain Filter (GBM4D) projection, to enhance dynamic PET image reconstruction.

METHODS

The sinogram was first transformed using the Anscombe method, then denoised using a combination of hard thresholding and Wiener filtering. Each denoising step involved guided block matching and grouping, collaborative filtering, and weighted averaging. The guided block matching was performed on accumulated PET sinograms to prevent mismatching due to low photon counts. The performance of the proposed denoising method (GBM4D) was compared to other methods such as wavelet, total variation, non-local means, and BM3D using computer simulations on the Shepp-Logan and digital brain phantoms. The denoising methods were also applied to real patient data for evaluation.

RESULTS

In all phantom studies, GBM4D outperformed other denoising methods in all time frames based on the structural similarity and peak signal-to-noise ratio. Moreover, GBM4D yielded the lowest root mean square error in the time-activity curve of all tissues and produced the highest image quality when applied to real patient data.

CONCLUSION

GBM4D demonstrates excellent denoising and edge-preserving capabilities, as validated through qualitative and quantitative assessments of both temporal and spatial denoising performance.

Noise2Void: unsupervised denoising of PET images

Objective:Elevated noise levels in positron emission tomography (PET) images lower image quality and quantitative accuracy and are a confounding factor for clinical interpretation. The objective of this paper is to develop a PET image denoising technique based on unsupervised deep learning.Significance:Recent advances in deep learning have ushered in a wide array of novel denoising techniques, several of which have been successfully adapted for PET image reconstruction and post-processing. The bulk of the deep learning research so far has focused on supervised learning schemes, which, for the image denoising problem, require paired noisy and noiseless/low-noise images. This requirement tends to limit the utility of these methods for medical applications as paired training datasets are not always available. Furthermore, to achieve the best-case performance of these methods, it is essential that the datasets for training and subsequent real-world application have consistent image characteristics (e.g. noise, resolution, etc), which is rarely the case for clinical data. To circumvent these challenges, it is critical to develop unsupervised techniques that obviate the need for paired training data.Approach:In this paper, we have adapted Noise2Void, a technique that relies on corrupt images alone for model training, for PET image denoising and assessed its performance using PET neuroimaging data. Noise2Void is an unsupervised approach that uses a blind-spot network design. It requires only a single noisy image as its input, and, therefore, is well-suited for clinical settings. During the training phase, a single noisy PET image serves as both the input and the target. Here we present a modified version of Noise2Void based on a transfer learning paradigm that involves group-level pretraining followed by individual fine-tuning. Furthermore, we investigate the impact of incorporating an anatomical image as a second input to the network.Main Results:We validated our denoising technique using simulation data based on the BrainWeb digital phantom. We show that Noise2Void with pretraining and/or anatomical guidance leads to higher peak signal-to-noise ratios than traditional denoising schemes such as Gaussian filtering, anatomically guided non-local means filtering, and block-matching and 4D filtering. We used the Noise2Noise denoising technique as an additional benchmark. For clinical validation, we applied this method to human brain imaging datasets. The clinical findings were consistent with the simulation results confirming the translational value of Noise2Void as a denoising tool.

A Multiprocessing Scheme for PET Image Pre-Screening, Noise Reduction, Segmentation and Lesion Partitioning

OBJECTIVE

Accurate segmentation and partitioning of lesions in PET images provide computer-aided procedures and doctors with parameters for tumour diagnosis, staging and prognosis. Currently, PET segmentation and lesion partitioning are manually measured by radiologists, which is time consuming and laborious, and tedious manual procedures might lead to inaccurate measurement results. Therefore, we designed a new automatic multiprocessing scheme for PET image pre-screening, noise reduction, segmentation and lesion partitioning in this study. PET image pre-screening can reduce the time cost of noise reduction, segmentation and lesion partitioning methods, and denoising can enhance both quantitative metrics and visual quality for better segmentation accuracy. For pre-screening, we propose a new differential activation filter (DAF) to screen the lesion images from whole-body scanning. For noise reduction, neural network inverse (NN inverse) as the inverse transformation of generalized Anscombe transformation (GAT), which does not depend on the distribution of residual noise, was presented to improve the SNR of images. For segmentation and lesion partitioning, definition density peak clustering (DDPC) was proposed to realize instance segmentation of lesion and normal tissue with unsupervised images, which helped reduce the cost of density calculation and completely deleted the cluster halo. The experimental results of clinical data demonstrate that our proposed methods have good results and better performance in noise reduction, segmentation and lesion partitioning compared with state-of-the-art methods.

PET image denoising using unsupervised deep learning

PURPOSE

Image quality of positron emission tomography (PET) is limited by various physical degradation factors. Our study aims to perform PET image denoising by utilizing prior information from the same patient. The proposed method is based on unsupervised deep learning, where no training pairs are needed.

METHODS

In this method, the prior high-quality image from the patient was employed as the network input and the noisy PET image itself was treated as the training label. Constrained by the network structure and the prior image input, the network was trained to learn the intrinsic structure information from the noisy image and output a restored PET image. To validate the performance of the proposed method, a computer simulation study based on the BrainWeb phantom was first performed. A 68Ga-PRGD2 PET/CT dataset containing 10 patients and a 18F-FDG PET/MR dataset containing 30 patients were later on used for clinical data evaluation. The Gaussian, non-local mean (NLM) using CT/MR image as priors, BM4D, and Deep Decoder methods were included as reference methods. The contrast-to-noise ratio (CNR) improvements were used to rank different methods based on Wilcoxon signed-rank test.

RESULTS

For the simulation study, contrast recovery coefficient (CRC) vs. standard deviation (STD) curves showed that the proposed method achieved the best performance regarding the bias-variance tradeoff. For the clinical PET/CT dataset, the proposed method achieved the highest CNR improvement ratio (53.35% ± 21.78%), compared with the Gaussian (12.64% ± 6.15%, P = 0.002), NLM guided by CT (24.35% ± 16.30%, P = 0.002), BM4D (38.31% ± 20.26%, P = 0.002), and Deep Decoder (41.67% ± 22.28%, P = 0.002) methods. For the clinical PET/MR dataset, the CNR improvement ratio of the proposed method achieved 46.80% ± 25.23%, higher than the Gaussian (18.16% ± 10.02%, P < 0.0001), NLM guided by MR (25.36% ± 19.48%, P < 0.0001), BM4D (37.02% ± 21.38%, P < 0.0001), and Deep Decoder (30.03% ± 20.64%, P < 0.0001) methods. Restored images for all the datasets demonstrate that the proposed method can effectively smooth out the noise while recovering image details.

CONCLUSION

The proposed unsupervised deep learning framework provides excellent image restoration effects, outperforming the Gaussian, NLM methods, BM4D, and Deep Decoder methods.

Joint solution for PET image segmentation, denoising, and partial volume correction

Segmentation, denoising, and partial volume correction (PVC) are three major processes in the quantification of uptake regions in post-reconstruction PET images. These problems are conventionally addressed by independent steps. In this study, we hypothesize that these three processes are dependent; therefore, jointly solving them can provide optimal support for quantification of the PET images. To achieve this, we utilize interactions among these processes when designing solutions for each challenge. We also demonstrate that segmentation can help in denoising and PVC by locally constraining the smoothness and correction criteria. For denoising, we adapt generalized Anscombe transformation to Gaussianize the multiplicative noise followed by a new adaptive smoothing algorithm called regional mean denoising. For PVC, we propose a volume consistency-based iterative voxel-based correction algorithm in which denoised and delineated PET images guide the correction process during each iteration precisely. For PET image segmentation, we use affinity propagation (AP)-based iterative clustering method that helps the integration of PVC and denoising algorithms into the delineation process. Qualitative and quantitative results, obtained from phantoms, clinical, and pre-clinical data, show that the proposed framework provides an improved and joint solution for segmentation, denoising, and partial volume correction.

Dynamic positron emission tomography image restoration via a kinetics-induced bilateral filter

Dynamic positron emission tomography (PET) imaging is a powerful tool that provides useful quantitative information on physiological and biochemical processes. However, low signal-to-noise ratio in short dynamic frames makes accurate kinetic parameter estimation from noisy voxel-wise time activity curves (TAC) a challenging task. To address this problem, several spatial filters have been investigated to reduce the noise of each frame with noticeable gains. These filters include the Gaussian filter, bilateral filter, and wavelet-based filter. These filters usually consider only the local properties of each frame without exploring potential kinetic information from entire frames. Thus, in this work, to improve PET parametric imaging accuracy, we present a kinetics-induced bilateral filter (KIBF) to reduce the noise of dynamic image frames by incorporating the similarity between the voxel-wise TACs using the framework of bilateral filter. The aim of the proposed KIBF algorithm is to reduce the noise in homogeneous areas while preserving the distinct kinetics of regions of interest. Experimental results on digital brain phantom and in vivo rat study with typical ¹⁸F-FDG kinetics have shown that the present KIBF algorithm can achieve notable gains over other existing algorithms in terms of quantitative accuracy measures and visual inspection.

Segmentation based denoising of PET images: an iterative approach via regional means and affinity propagation

Delineation and noise removal play a significant role in clinical quantification of PET images. Conventionally, these two tasks are considered independent, however, denoising can improve the performance of boundary delineation by enhancing SNR while preserving the structural continuity of local regions. On the other hand, we postulate that segmentation can help denoising process by constraining the smoothing criteria locally. Herein, we present a novel iterative approach for simultaneous PET image denoising and segmentation. The proposed algorithm uses generalized Anscombe transformation priori to non-local means based noise removal scheme and affinity propagation based delineation. For nonlocal means denoising, we propose a new regional means approach where we automatically and efficiently extract the appropriate subset of the image voxels by incorporating the class information from affinity propagation based segmentation. PET images after denoising are further utilized for refinement of the segmentation in an iterative manner. Qualitative and quantitative results demonstrate that the proposed framework successfully removes the noise from PET images while preserving the structures, and improves the segmentation accuracy.

The 2D Hotelling filter - a quantitative noise-reducing principal-component filter for dynamic PET data, with applications in patient dose reduction

Background

In this paper we apply the principal-component analysis filter (Hotelling filter) to reduce noise from dynamic positron-emission tomography (PET) patient data, for a number of different radio-tracer molecules. We furthermore show how preprocessing images with this filter improves parametric images created from such dynamic sequence.

We use zero-mean unit variance normalization, prior to performing a Hotelling filter on the slices of a dynamic time-series. The Scree-plot technique was used to determine which principal components to be rejected in the filter process. This filter was applied to [¹¹C]-acetate on heart and head-neck tumors, [¹⁸F]-FDG on liver tumors and brain, and [¹¹C]-Raclopride on brain. Simulations of blood and tissue regions with noise properties matched to real PET data, was used to analyze how quantitation and resolution is affected by the Hotelling filter. Summing varying parts of a 90-frame [¹⁸F]-FDG brain scan, we created 9-frame dynamic scans with image statistics comparable to 20 MBq, 60 MBq and 200 MBq injected activity. Hotelling filter performed on slices (2D) and on volumes (3D) were compared.

Results

The 2D Hotelling filter reduces noise in the tissue uptake drastically, so that it becomes simple to manually pick out regions-of-interest from noisy data. 2D Hotelling filter introduces less bias than 3D Hotelling filter in focal Raclopride uptake. Simulations show that the Hotelling filter is sensitive to typical blood peak in PET prior to tissue uptake have commenced, introducing a negative bias in early tissue uptake. Quantitation on real dynamic data is reliable. Two examples clearly show that pre-filtering the dynamic sequence with the Hotelling filter prior to Patlak-slope calculations gives clearly improved parametric image quality. We also show that a dramatic dose reduction can be achieved for Patlak slope images without changing image quality or quantitation.

Conclusions

The 2D Hotelling-filtering of dynamic PET data is a computer-efficient method that gives visually improved differentiation of different tissues, which we have observed improve manual or automated region-of-interest delineation of dynamic data. Parametric Patlak images on Hotelling-filtered data display improved clarity, compared to non-filtered Patlak slope images without measurable loss of quantitation, and allow a dramatic decrease in patient injected dose.